Rituximab Induces an Interferon Gene Expression Signature in Patients with CLL.

Rituximab, a monoclonal anti-CD20 antibody, is used to treat Chronic Lymphocytic Leukemia (CLL) in combination with fludarabine. Rituximab is thought to deplete B-cells through antibody-dependent cellular cytotoxicity, complement-dependent cytotoxicity, and possibly signaling for apoptosis. Whether or not signaling by rituximab contributes to its clinical efficacy and can sensitize the malignant cells to chemotherapy is controversial. To investigate if rituximab can induce a specific gene expression signature, we used genomic-scale gene expression profiling (Affymetrix HU133A 2.0 arrays) of B-cells from CLL patients receiving their first rituximab infusion. During the infusion, patients experienced a cytokine release syndrome (fever, chills, and hypotension) that led to interruption and symptomatic treatment in most; however, all patients were able to finish the treatment. Absolute lymphocyte counts decreased on average by 50% over this initial 24h period. We analyzed CD19+ selected CLL cells from eight patients obtained pre and 6 and 24 hours after the start of rituximab. A one-way ANOVA test was used to identify genes up- or down-regulated with a false discovery rate (FDR = number of expected chance findings / number of observations) of <10%. We identified 80 genes with at least 1.5× higher expression at 6h versus 0h including many interferon (IFN)-regulated genes like IRF1, IFITM1, STAT1, JAK3 and several apoptosis related genes such as FAS and Caspase 8. The majority of these genes were at least 2-fold up-regulated at 6 hours, but most returned to pre-treatment levels by 24 hours. Thus, rituximab induced a transient gene expression signature that correlated with the cytokine release syndrome during the infusion. To determine whether or not this IFN signature was caused directly by rituximab signaling or indirectly by cytokines released during the infusion, we compared rituximab and IFN gamma effects on CLL cells in-vitro. Both rituximab (10ug/ml with cross-linking) and IFN-gamma (1000U/ml) induced FAS (CD95) expression in CLL cells measured by flow cytometry. CD95 expression was low on untreated CLL cells, at 6 hours, up-regulation of CD95 expression with rituximab was stronger than with IFN, while at 24 hours, IFN treated cells showed slightly higher CD95 expression. Next, we investigated whether rituximab is able to activate STAT1, the main transcription factor regulating IFN target genes. IFN gamma induced rapid phosphorylation of STAT1 in CLL cells, but rituximab did not. However, we observed phosphorylation of ERK in response to rituximab as has been reported by others. After in-vitro stimulation with rituximab and IFN-gamma, IRF-1 and STAT-1 were up-regulated at 2 and 6 hours as measured by real time PCR, albeit with a stronger response after IFN. We conclude that rituximab is associated with a specific gene expression signature in CLL patients that is characterized by IFN response genes. At this point, we cannot rule out that this signature is contributed in part by cytokines released during the rituximab infusion. However, the rapid up-regulation of CD95, STAT1, and IRF1 under controlled in-vitro conditions is consistent with a direct effect of rituximab. Ongoing studies aim to better characterize rituximab signaling in CLL and to determine whether this can contribute to apoptosis or sensitize the leukemic cells to chemotherapy.